Data Mining

My excuse for the late blog post this week (and last) Not that I’ve gotten past page 25, but so far a good read, espcecially for a government document. It’s been three days since I wrote that last sentence, so I’m going to publish today, finished or not.


When I started teaching, 15 years ago, No Child Left Behind, NCLB, was not a thing. Common Core was not a thing, I don’t think most states even had standards. Fifteen years later and NCLB is over, replaced with Every Child Succeeds, but the damage perhaps has been done.


Most educators I know have been pretty negative on the whole NLCB thing. As for me, I thought it was a joke from the beginning. Who would ever expect 100% achievement? Haven’t you ever heard of SPED, ELL, ED, children? Not to mention the obvious inequities in buildings or supports not controlled by schools. I mean really 100%, not going to happen, ever! What I did like was the conversation on accountability. I would like to think we could have started that conversation without threatening to destroy public education, but maybe not.


The core problem, as I see it, is that accountability has never been defined in a way that actually improves education. I think teachers need to be accountable to parents, after all it’s their children we are educating. When asked most parents will usually say something to the effect of, my school is good, it’s those other schools that suck. Those few parents who will look at their own school and say this school sucks will usually say, but here are good teachers in the school. Finally, everyone, even the pope, will agree good schools, and teachers in particular, are underfunded, yet try to get a local school referendum to pass in the US. Why is it we demand accountability without furnishing the tools to succeed?


Anyway my point is not to whine about the state of our educational system in the United States it could be better but it will not change because I wrote a blog post. What I would really like to talk about today is data.

What is data?

This whole NCLB thing heralded a new emphasis on data. We can’t trust principals, those experts of teaching so named because they are the principal teacher in a school. We can’t trust parents, because they are intimidated by the degree carrying hooligans. And we certainly can’t trust those unions, they are the cause of all this trouble. We have to use hard fast, data real numbers that don’t lie. (Please don’t quote Mark Twain and statistics right now).


I think education can be improved with the use of data but we’re not having a conversation on how to create and mine that data. What we’re trying to do is collect a ton of data on everything without getting too deep into the question of what makes quality data.


Evidently, if data is to tell the truth it must be a number and that number must be pulled from so sort of test. If we want to know anything about education, we should ignore quotes attributed to Confucius. In this manner everything that is done in education must be subject to some sort of double blind or A B test. If we are to intervene with students that intervention must have been researched and found acceptable. And when we are not using an approved and researched curriculum (and sometimes when we are), we should be also be performing our own research.

How do we find good data for our classroom?

Teachers are often being asked to write and or use a common formative assessments and then use that data to inform education. We are then to compare our results from a common assessment with the results of the teacher down the hall. If they are doing a better job we should consider teaching in the same manner (because it couldn’t have anything to do with the students or anything like that). Like the PARCC test but on a smaller scale.


Some questions I have on this whole thing. Does a multiple choice test with a hundred questions give me good data? Can a three question exit slip give me good data? Can a common formative assessment really lead to quality data? What is authentic assessment? Will authentic data in my classroom be the same as authentic data in the classroom next door? If a student can explain how to solve a problem, but can’t do that on paper, has s/he mastered the standard? If data is so important why isn’t psychometrics a class in teacher education? Why don’t we have a psychometrician in each school or at least each district?


Good classroom data won’t come from some administrative committee outside of the classroom creating assessments. Creating formative assessments has to be done by the teachers who are teaching that subject. Not teachers of the subject, but teachers actually teaching right now. Some sort of common summative assessment might be a different beast. Do teachers have discussions on what constitutes a quality question that provides quality data? Do they have that kind of time? How do we standardize the data collection from authentic assessments? If I use problem based learning and she uses direct instruction can we collect the same data? Can we compare data?

Teacher education

Are teachers taught how to create a vehicle for collecting quality data? Why do we always seem to default to a test? Do we know what makes a good question? When learning to teach we study child development. We have an idea of how a students in our age range will generally behave, and act. We learn pretty quick about the neighborhood in which we teach. What resources will a student have at home. How much support they will get when doing homework. That sort of thing, but how does that translate into data? More importantly how does that translate into data we can use in the classroom to inform instruction?


I haven’t been in a teacher education program for a long time. When I was there we had a couple of classes on data collection and action research, but I don’t remember a big emphasis on quality data collection methods in the classroom, just action research for our own purposes. It was a graduate program so our data collection methods were concentrated on data we would use for our thesis paper. How similar or different should that be to data collection in our classroom. Right now I have to say the data collection we do for state and federal levels isn’t really all that high quality.


I’m not even talking testing scandals in Washington D.C. or Atlanta. I’m just talking about quality data, the basics. Sure we are testing 90+% of the kids, but are the kids actually putting in an effort. Obviously not, because  a lot of schools have these big expensive pep rally, assemblies, reward parties, motivate students to put effort into the test. Quite honestly, I think, we would be better off using real world modeling to choose a motivated representative sample of the population and use that to determine if education is working. The only problem is that we couldn’t do representative sampling for individual teachers, or even schools. Of course the data we use now is suspect for individual teachers and most schools as well, but it seems we aren’t willing to admit that just yet.

Data and learning

What do we do with the data? Brain based education is popular now. Teachers, at least some, are designing lessons based on engaging the whole child. No more drill and kill until education is beat into a child, instead we create multifaceted lessons designed to create an entry point for each child and all of them lead to the same end. In the classroom teachers should formatively assess students and use that information to change instruction for the better.  Subject teams create common formative assessments to see if one teacher is teaching better than another. Buildings should compare those common formative assessments within a district. So on and so forth until we know how to best educate each and every child in the best way.


Well that’s the dream. the reality is depending on the school, or even the classroom, I might spend the majority of my time keeping order, or reteaching, or back filling knowledge, or feeding the hungry, or sharing emotions. Data collection can easily get lost. I forget, I compromise, the well crafted assessment is scrapped for an interesting rabbit hole.


My point is this, we don’t talk enough about what it means to collect quality data in the classroom. We don’t talk enough about where and how it informs instruction. It’s there and some teachers are really good at it (not me), but that isn’t what I mean. I mean that is not the central focus of teacher education, but it seems to be the central focus of evaluation. So which one will change?  


Loading Facebook Comments ...

Leave a Reply